DESCRIPTION: Identification of early markers for negative health outcomes is critical to maintaining optimal infant development and for developing successful health interventions, particularly for very low birth weight (VLBW, birth weight less than 1,500 g) preterm male infants who are more prone to health-related problems than female infants. General gender-difference theories and the extreme male brain theories, as well as our preliminary studies suggest that testosterone and cortisol levels are positively related to each other in infants and are negatively associated with infant health outcomes in early infancy and with developmental outcomes between ages 1 to 3. We plan to confirm that testosterone rather than cortisol is a more reliable marker/predictor of complications affecting infants' health outcomes, mother-infant interactions, and infant cognitive/motor/ language developmental outcomes; and that male infants exhibit a higher sensitivity to testosterone levels than female infants as evidenced by these outcomes. This study involves a comparative and longitudinal research design aimed at describing gender differences measured by testosterone and cortisol levels with (1) health outcomes of neonates and infants, (2) quality of mother-infant interactions, and (3) infant cognitive/motor/language development after adjusting for maternal psychological state, infant temperament, and characteristics of mothers. One hundred ninety mother-VLBW preterm infant pairs will be recruited from the neonatal intensive care unit (NICU) at the University of Alabama at Birmingham affiliated Women and Infants Center. Demographic and health data of the pairs will be collected at the NICU and updated at the newborn follow-up clinic where the infants are scheduled for three follow-up visits (at 6, 12, and 24 months corrected age for prematurity) following discharge. During these visits, mother-infant interactions will be videotaped for 15 min and infant cognitive/motor/language developmental outcomes will be collected using age-appropriate measurements. Concurrent and repeated measurement of testosterone and cortisol levels both in infants and in mothers will be conducted through infancy and early childhood (at birth, 40 weeks postmenstrual age and 6 , 12 , and 24 months corrected age). Exploratory data analysis will first be performed to gain insight into the collected data. Generalized linear models will be used to model independent data with one time collection while the generalized estimating equation and the generalized linear mixed-effects models are standard approaches for modeling correlated longitudinal data.